---
output: 
  flexdashboard::flex_dashboard:
    theme: bootstrap
    source: embed
---

```{r setup, include=FALSE}
library(flexdashboard)
```


Introduction {data-icon="fa-book-open"}
=====================================  


Tidying Data {data-icon="fa-table"}
=====================================  

### **Loading Files From Kaggle**

```{r echo=TRUE, message=FALSE, warning=FALSE}
#------LOAD DATA------------
library(tidyverse)
library(readxl)
library(leaflet)
library(flexdashboard)
library(gridExtra)
library(broom)
library(plotly)

#------LOAD FILES---------
CCodesfile="Country-Code.xlsx"
Zomatofile="zomato.xlsx"
CurrencyFile="CurrencyRates.xlsx"

CCodes=read_excel(CCodesfile) #importing country code file
zomato=read_excel(Zomatofile) #importing main zomato file
Currency=read_excel(CurrencyFile) #importing currency file

```

> The original zomato data set contained varying currencies per country, making it difficult to compare each restaurant fairly. To approach this problem, we created a seperate currency table containing USD conversion rates per country.

###Data Wrangling

```{r echo=TRUE, warning=FALSE}
#------ GET RID OF SOME COLUMNS & ADD NEW ONES -------------
zomato=zomato %>% select (-c(Rating_Color,Switch_To_Order_Menu,Address,Locality_Verbose,Currency))
#we remove currency column as it is incorrect and contains too many unique symbols

#add in new columns: Country Names according to Country Code & proper currency names with conversion rates
zomato = zomato %>% left_join(CCodes) 
zomato = zomato %>% left_join(Currency) 

#add in new column: USD Cost (for equal comparison)
zomato = zomato %>% mutate(Avg_Cost_USD = Average_Cost_For_Two*`Conversion Rate (USD)`) 
#Multiplied Currency Column by it's conversion rate

#replacing Cuisines col. with Principal_Cuisines (the first category of cuisines for every country row)
zomato= zomato %>% separate(Cuisines,into=c("Principal_Cuisines"),sep=',') #store "primary" cuisine types into new column and replace old with this

#Make sure rating categories are ordered:
zomato = zomato %>% mutate(Rating_Text=ordered(Rating_Text,levels=c("Excellent","Very Good","Good","Average","Poor")))
```

> Although the data is presented pre-cleaned into appropriate columns, there still remains some level of manipulation that needs to be done. We begin by removing unnecessary columns and adding in necessary ones.
> We add in Countries by country code via the country code file provided by kaggle
> Then we add in our own currency conversion file, allowing us to convert average cost for two into the USD amount and storing it in it's own column.
> There were many cuisines listed per restaurant, so to simplify things for analytic purposes, we extract the first cuisine listed by each restaurant.


###Removing Missing Values
```{r echo=TRUE, warning=FALSE}
#---------REMOVE MISSING VALUES----------
#we notice that there are several "0" rated stores for even expensive places so we decided to remove no rating stores
zomato[zomato == 0] = NA #remove values that have zero
zomato[zomato == "Not rated"] = NA #remove unrated values
zomato = zomato %>% filter(!is.na(Aggregate_Rating))
```


> Since we will focusing on the ratings of restaurants, we remove restaurants with missing values.


Exploratory Analysis {data-icon="fa-chart-pie" .storyboard}
=====================================  

### Distribution of Quantative Variables

```{r}
#look at distributions of variables
ggplotly(
  ggplot(zomato,aes(Aggregate_Rating))+geom_histogram()
  )
#almost a normal distriubtion
```

***
COMMENTS INSERT HERE

### Cost Distribution 
```{r}
ggplotly(
  ggplot(zomato,aes(Avg_Cost_USD))+geom_histogram() #skewed to the right
)
```

### Vote Distribution
```{r}
ggplotly(
  ggplot(zomato,aes(Votes))+geom_histogram() #skewed right
)
```

### Distribution of Countries
```{r echo=FALSE}
library(RColorBrewer)
#themes for plot
plot_themes=theme(plot.title=element_text(family="Palatino", face="bold.italic", size=15),legend.title = element_text(face = "italic", family = "Palatino"), 
                  legend.text = element_text(face = "bold.italic",family = "Palatino"), 
                  axis.title.y = element_text(family = "Palatino", face="italic",size = (10)),
                  axis.text.y.left = element_text(family = "Palatino", colour = "darkgrey", size = (10)),
                  axis.text.x.bottom = element_blank(),
                  axis.ticks.x = element_blank())

getPalette= colorRampPalette(brewer.pal(9,"RdBu"))

#Closer look into Country
Country_Frequency=ggplot(zomato,aes(Country))+geom_bar(aes(fill=Country)) + 
  scale_fill_manual(values = getPalette(15))

#printing with themes and titles 
ggplotly(
  Country_Frequency + plot_themes + labs(title="Country Frequency",y="Country Count",x="Countries",fill="Countries")
)
```


###
First Stage Analysis {data-icon="fa-chart-bar"}
=====================================  



Second Stage Analysis {data-icon="fa-chart-line"}
=====================================  


Conclusions {data-icon="fa-book"}
===================================== 


Discussion {data-icon="fa-comment-alt"}
=====================================  


References {data-icon="fa-list-alt"}
=====================================